adhd-output-style

Formats AI responses into short numbered steps with progress markers for limited working memory readers.

1.1k|107|Updated Jul 9, 2025
One-click install
npx skills add https://github.com/fcakyon/claude-codex-settings --skill adhd-output-style
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: adhd-output-style
Source: https://github.com/fcakyon/claude-codex-settings/tree/main/plugins/adhd-output-style/skills/adhd-output-style
Command: npx skills add https://github.com/fcakyon/claude-codex-settings --skill adhd-output-style

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Long, unstructured AI responses overwhelm readers with limited working memory, burying the actionable answer under preambles, recaps, and vague language. This Skill enforces a response format that surfaces the answer first and keeps every step scannable.

Core Features & Use Cases

  • Action-First Structure: Opens with the answer or next step, breaks work into numbered one-action steps, and ends with a single next action under two minutes.
  • Progress Restatement: Restates progress each turn (e.g., "step 3 of 5"), caps lists at five items, and uses concrete time estimates instead of vague ones.
  • Educational Insight Blocks: Adds a short codebase-specific insight block before and after code, keeping depth separate from the terse main answer.
  • Use Case: A developer with ADHD asks for help debugging a build failure. The response starts with the fix, lists three numbered steps with time estimates, states what now works, and ends with one quick next action.

Quick Start

Ask the assistant to use ADHD output style with short numbered steps for the current task.

Frequently Asked Questions about adhd-output-style

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I make AI responses shorter and easier to follow?

Invoke the adhd-output-style skill, which opens with the answer, breaks work into numbered one-action steps, and cuts preambles and recaps. Each response ends with a single next action that takes under two minutes.

What is an ADHD-friendly output format for AI assistants?

It is a response structure that leads with the actionable step, uses numbered lists capped at five items, restates progress each turn, and gives concrete time estimates. Errors are described factually with cause and fix, without alarmed language.

Does this output style still explain code and teach concepts?

Yes. Educational content goes into a separate insight block with two to three codebase-specific points placed before and after code. The main answer stays terse while depth lives in the insight block.

When does the skill give full walkthroughs instead of short steps?

Full walkthroughs are given when explicitly asked. The skill also confirms before destructive actions, pauses with a diagnostic question after repeated failed debugging, and asks one clarifying question on genuine ambiguity.

Can I use this formatting style with Claude Code or Codex CLI?

Yes. The skill ships as a plugin in the claude-codex-settings repository and installs into Claude Code, Codex CLI, and Cursor via the plugin marketplace. Once active, it applies to all interactions in the current task.